Improving the learning effectiveness of educational videos

This paper presents a tool that improves the learning effectiveness of educational/instructional videos by automatically generating an hyperlinked multimedia structure created by 1) analyzing the audio stream to extract a textual description; 2) carrying out semantic analysis to identify the main subjects addressed; 3) identifying keywords and keyframes, and creating a navigational structure that allows the viewers to consume the content in a non-linear and multimodal manner and jump directly into those parts they are mostly interested to watch. The tool performs transformation and automatic curation of any video, allowing the viewers to explore its content in a non-linear form. It can be applied to any type of videos, but it is of particular value in the educational sector, where videos are frequently viewed in parts, and revisited frequently to recall specific aspects of a lecture/topic.